We completed genotyping subjects with both the Affymetrix 100,000 SNP chip and the 1 million SNP chip technologies. Phase 1 was designed to detect potential associations with young-onset type 2 diabetes, by genotyping 300 early-onset type 2 diabetes subjects (onset age<25 yrs) and 329 non-diabetic controls (age >45 yrs), and 271 additional subjects who were diabetic and non-diabetic siblings of the selected subjects. Associations with diabetes were calculated using both a case/control analysis (N= 629) and a within-family analysis (482 siblings from 169 sibships), and SNPs that had the strongest association for the combined associations were prioritized. Phase 2 of the GWA was designed to detect associations with pre-diabetic traits (% body fat, insulin action as measured by the hyperinsulinemic euglycemic clamp technique, and the acute insulin response to an intravenous bolus of glucose). Six hundred non-diabetic subjects who had been metabolically phenotyped for these predictors of diabetes were genotyped. Measures of BMI were available on all samples from Phase 1 and Phase 2. SNPs that provided the best associations for diabetes and/or a pre-diabetic trait (including BMI) were selected for additional genotyping in a population-based sample of 3501 full-heritage Pima Indians. This genotyping utilized a new high throughput technology (Bead Express). We recently completed genotyping the best signals from our genome-wide association analysis in a sample of 3501 full hertiage Pima Indians and are currently replicating the best SNPs from this sample in a second population-based sample of 3784 mixed heritage Native Americans. To date the strongest assocations in the combined samples for diabetes are with SNPs in DNER (P= 1 x 10-8) and with KCNQ1 (P= 5 x 10-9). The strongest associations in the combined samples for BMI are with SNPs in BRD2, HEATR5B, UBE2E, GSTA5, NOVA1, FTO, MAP2K3, and CYB5A (all P between 10-6 and 10-7). Studies are ongoing to identify the casual variant that underlies each of these associations.